Universal Laws and Architectures for Complex Networks
There are universal tradeoffs in distributed control systems and universal architectures for managing them, illustrated here by standard computer software/hardware stack, the analogous stack in the bacterial biosphere, and a cartoon of a conjectured nervous system “stack.”
There are universal tradeoffs in distributed control systems and universal architectures for managing them. Perhaps most familiar is the software/hardware stack in computer systems, where application software provides the most flexibility, but the highest speed responses in any technology requires specialized and inflexible hardware implementations. The operating system (OS) and various utilities and CAD tools facilitate the implementation of suitably placed control system functionality. A shared OS within a network facilitates swapping of apps and hardware. Similarly, the bacterial biosphere has a shared architecture with translation replacing the OS, and swapping genes by horizontal gene transfer replacing “horizontal app transfer”? We conjecture that a similar architecture is involved in both learning and evolution in the human nervous system, but the “OS” component is less clear, and will be the focus of our research.
